The present invention relates to an information search terminal apparatus (also referred to simply as terminal) in an information search/retrieval system for searching from a database storing document data and the like a desired document as a whole by using a designated character string as a search term. More particularly, the invention is concerned with information search terminal and system which can afford improved manipulatability to the user.
In the document search or retrieval systems known heretofore, search can be performed only through a continuous flow of processings, starting from the inputting of search conditions or queries till the display of a document resulting from the search, as is disclosed in JP-A-62-105270. In other words, the document search had to be performed through a series of sequential processings of inputting the search queries, display of search results in the from of a list and then display of the document data. More specifically, by selecting a document (ID) number from the list of the search results, contents of the corresponding document can be displayed. However, in order to designate another document the content of which is to be displayed, manipulation for displaying once again the list of search results is required. This means that the flow of processings performed by the user will necessarily be interrupted and the search process has to be executed once again from the beginning, providing thus a cause for degradation in efficiency of the document search operation.
There is proposed in JP-A-64-35662 a document search system which is capable of displaying simultaneously a plurality of samples of documents as searched, as an approach for improving the efficiency of the document search. This system is however disadvantageous in that once one document has been selected from a plurality of sample images, all the processes carried out till than are canceled, as a result of which the sample image list can no more be restored even when it becomes necessary to select again another document. It goes without saying that such processing which cancels the processes executed in the past will remarkably degrade the efficiency of document search operation.
Further, in the document search system disclosed in JP-A-62-11932, there is proposed a technique for expanding from a search term designated by the user other character strings or terms having a same meaning as the designated search term before searching a database. However, in this known search system, no information is available for the user to know which of the terms inclusive of the expanded one is used as the search term in the search processing performed currently, to discomfortableness of the operator. Accordingly, the search should be performed after having confirmed the results of expansion of the search term as specified or designated. However, necessity of confirming the expansions of the search term upon every search will impose troublesome procedure or burden to the user, presenting thus another problem.
Additionally, it is noted that in the course of the search, there often arises a need for screening a set of documents resulting from a preceding search in the light of additional or new queries or conditions. By way of example, let's assume that one million documents registered in a document database can be reduced to ten thousand documents through a screening search process and that the ten thousand documents be reduced to about one hundred documents through a second screening process. In carrying out such screening, there are conceivable two modes, i.e., a mode in which the search is performed on the basis of a set of results obtained from an immediately preceding search and a mode in which the search is performed on the basis of a specific document set by applying various query conditions in a trail-and-error manner. For convenience of description, the first mentioned search mode will be referred to as "hierarchy search" with the second mode being termed "universe mode". In order to realize the screening search mentioned above, there is required means for informing the user whether the search performed currently is for all the documents stored in the database or a screening search is performed on a base set (generatrix set) including the result of the preceding search or in which mode of hierarchy or universe search and search is currently performed.